More infections in poorly ventilated classrooms

Cooperation for Grisons schools: Patrik Degiacomi, Head of the Department of Education, Society, Culture, Empa researcher Hossein Gorji and Martin Bühler, Head of the Cantonal Command Staff (from left to right). Image: Südostschweiz/Olivia Aebli-Item

Air quality has been measured in 150 classrooms of 59 elementary school in Graubünden since October 2021. Two CO₂ sensors were installed in each room. A first evaluation, with data sets collected between 1 and 19 November 2021, is now available. A statistical correlation between the number of positive cases in the weekly school tests and the measured CO₂ concentration in the corresponding classrooms could be established: Significantly more individuals were infected with SARS-CoV-2 in poorly ventilated classrooms.

Moreover, the air quality measurements show that the majority of the classrooms under investigation have elevated levels of CO₂. In 60 percent of the classrooms, CO₂ levels exceeded 2000 parts per million (ppm). Elevated CO₂ concentrations can impair cognitive performance and thus learning ability.

Long-term measurement: A sensor for carbon dioxide and other data in a classroom in Chur. Image: Canton of Grisons / Mattias Nutt Photography

How CO₂ concentrations in indoor spaces correlates with the transmission of air-borne diseases had not yet been empirically established. In the pilot phase of the current study, the Graubünden Health Department therefore had sensors installed in around 150 classrooms in 59 schools throughout the canton to record CO₂ levels – two sensors in each room to back up the measurements. They continuously recorded CO₂ concentrations and other parameters such as temperature and air pressure, while weekly Corona mass tests were conducted on students and teachers.

The preliminary results are based on data collected between 1 and 19 November. The number of infections per 100 students, known as prevalence, correlated with the CO₂ concentration in classrooms. In high-risk schools, it often exceeded 3,000 parts per million (ppm) – a level significantly higher than the recommended level of no more than 1,000 ppm.

However, statistics in and by itself is not sufficient to accurately describe the correlation between infection incidence and CO₂ levels. In addition, the analysis of the Empa researchers is subject to certain limitations. For instance, besides the "cumulative" CO₂ content, i.e. the summed amount of the gas, there could well be other indicators of the exposure risk to infectious aerosols.

Therefore, the experts will analyze other variables from their CO₂ time series, such as the temporal variation of the concentration and temperature changes – with the goal of finding dependencies between infection prevalence and recorded CO₂ levels.